Retail Business Convenience Segmentation using Clustering and Data Visualization

Abstract

The conventional approach to launching a business is to research and gather data regarding the past performance of rival businesses unless they were profitable or unsuccessful. Innovation is the ethos of the modern day, as everyone is engaged in a struggle to outperform one another. The objective of our suggested research is to create knowledge that will be helpful to aspiring business owners and small companies that are losing money. Our main aim is to assist small-scale manufacturers in becoming successful marketers. In return for the dataset, which must be provided as input, we will provide them with clear instructions on how to start a profitable business and recover from their loss. In order to analyse data more effectively, our planned work will segment clients based on stock input, weekly updates of stocks sold, and waste products. In this work, two different clustering techniques (k-Means and hierarchical) are used to classify the products into subsets, and their respective results are compared. Data will be segmented using clustering algorithms, allowing for much more focused production of the final result.